Improved synthesis of a trisphosphine ligand and crystallographic characterization of the ligand and nickel thiocyanate complex

Article abstract of DOI:10.1016/j.poly.2012.11.028

Improved syntheses of o-C₆H₄(PEt₂)(X) (X = F, Br, I) and the trisphosphine PhP[o-C₆H₄(PEt ₂)]₂ have been developed and are applicable to a wide variety of phosphine substituents.

Full text of DOI:10.1016/j.poly.2012.11.028

Polyhedron 58 (2013) 171–178
Contents lists available at SciVerse ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Improved synthesis of a trisphosphine ligand and crystallographic
characterization of the ligand and nickel thiocyanate complex
William J. Schreiter, Alexandre R. Monteil, Marc A. Peterson, Gregory T. McCandless,
⇑
Frank R. Fronczek, George G. Stanley
Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803-1804, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 3 December 2012
Improved syntheses of o-C6H4(PEt2)(X) (X = F, Br, I) and the trisphosphine PhP[o-C6H4(PEt2)]2 have been
developed and are applicable to a wide variety of phosphine substituents. The trisphosphine, PhP(o-Ph–
PEt2)2, has been fully characterized spectroscopically and, for the first time, crystallographically. Reaction
of the P3 ligand with one equivalent of NiCl2 or Ni(BF4)2 followed by 2.5 equivalents of KNCS produces the
Keywords:
monometallic complexes Ni(X)2[j
3-PhP(o-Ph–PEt2)2], (X = Cl or NCS). The nickel thiocyanate complex has
Phosphines
Ligand synthesis
Nickel
been characterized via a crystal structure. The nickel center has a square pyramidal structure with one of
ꢀ
the NCS ligands occupying the apical coordination site. The Ni complex crystallizes in the P1 space group
with an unusual six molecules in the asymmetric unit (Z0 = 6), and twelve in the unit cell.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
obtained using Pd-catalyzed coupling route (R = Ph, X = Br) [7] and
better solvents when using lithium reagents (R = Me, X = F) [8]. The
The strongly chelating trisphosphine, PhP(o-C6H4–PR2)2,
1
yields on preparing the trisphosphine ligand 1 from the appropri-
ate precursor 2 have also been quite low with yields of 16%
(R = Et) [1] and 35% (R = Ph) [2].
We report here improved and straightforward syntheses for 1
(R = Et) and especially 2 (R = Et, X = F, Br, I) along with the crystal-
lographic characterization of 1 and the nickel complex formed
from the reaction of Ni(NCS)2 and 1. This was part of our efforts
to synthesize a new, far more strongly chelating and bridging tet-
raphosphine ligand for catalytic studies, which will be reported
elsewhere.
(R = Et), was first prepared by Hart in 1960 [1]. Hartley, Venanzi
and coworkers reported the synthesis of the more commonly used
phenyl-substituted PhP(o-C6H4–PPh2)2, ligand in 1963 [2]. Since
then this P3 ligand, mainly in the form of the R = Ph, or sometimes
Me substituted versions, has been used to prepare a variety of
monometallic transition metal complexes and occasionally multi-
metallic systems [3–6]. The difficulty and low yields in synthesiz-
ing this ligand, especially the alkyl-substituted versions, has
limited its use in transition metal coordination chemistry.
2. Experimental
All manipulations were carried out under an inert atmosphere
of nitrogen and in oven-dried glassware using a Vacuum Atmo-
spheres Company glovebox or by using standard Schlenk tech-
niques unless noted otherwise. Tetrahydrofuran (THF), diethyl
ether, CH2Cl2, hexane, and N,N-dimethylformamide (DMF) were
obtained dry from Aldrich and processed through a Innovative
Technology Inc. PURESOLV™ solvent purification system under in-
ert atmosphere (N2). All other solvents were obtained anhydrous
from Aldrich and were used as received. Deionized water was
degassed by purging with nitrogen gas for 30 min prior to use.
Reagents and starting materials were obtained from commer-
cial suppliers and used as received, except for phenylphosphine
[9], which was prepared according to literature methods. Isopro-
pylmagnesium bromide was produced from magnesium turnings
The key to the synthesis of 1 is to improve the yield of the com-
monly used precursor 2. Early preparations 2 have reported yields
of: 21% (R = Cl, X = Br), 36% (R = Ph, X = Br) [2], 26% (R = Et, X = Br),
and 50% (R = Et, X = Cl) [1]. More recently, higher yields have been
⇑
Corresponding author.
E-mail address: gstanley@lsu.edu (G.G. Stanley).
0277-5387/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.poly.2012.11.028

172
W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
and isopropyl bromide in THF. Other organolithium and Grignard
reagents were obtained as commercial solutions from Aldrich
and were titrated over salicylaldehyde phenylhydrazone [10]
immediately preceding their use. Ni(BF4)2ꢀ6H2O and NiCl2ꢀ6H2O
were obtained from Strem Chemicals and used as received.
1H NMR spectra were recorded on a Bruker DPX-250, an ARX-
300, or a DPX-400 spectrometer with chloroform (7.26 ppm), ben-
zene (7.15 ppm) or dichloromethane (5.32 ppm) as the internal
standard. The 13C NMR spectra were recorded on the same instru-
ments with chloroform (77.0 ppm), or benzene (128.02 ppm) as
the internal standard. The 31P NMR spectra were recorded on the
same instruments with 85% phosphoric acid (0.0 ppm) as the
external reference. Multiplicity is reported as follows: s = singlet,
d = doublet, t = triplet, q = quartet, m = multiplet, om = overlapping
multiplets, dd = doublet of doublets, dt = doublet of triplets,
dq = doublet of quartets, td = triplet of doublets, and br = broad.
The multiplicity reported in the 13C{1H} spectra refers to the
31P–13C coupling.
2.3. Preparation of 1-(diethylphosphino)-2-fluorobenzene, 2f
Compound 2f was prepared by using the procedure described
for 2i. The fluoro product is not light sensitive. The reaction was
performed on a 0.171 mol scale using 1-bromo-2-fluorobenzene,
and afforded the production of 2f in 70–75% isolated yield
(23.0 g): bp = 77–80 °C (0.3 Torr).
31P{1H} NMR (101.2 MHz, d ppm, C6D6): ꢁ21.6 (d, JP,F = 29.8 Hz).
1H NMR (250 MHz, d ppm, C6D6): 7.2 (br m, 1H), 6.8 (m, 3H), 1.6
(m, 4H), 0.9 (m, J = 7.3 Hz, JP,H = 7.7 Hz, 6H). 13C{1H} NMR
(62.8 MHz, d ppm, C6D6): 133.8 (dd, J = 5.7 Hz, J = 13.4 Hz), 130.6
(d, J = 7.7 Hz), 124.2 (t, J = 3.8 Hz), 115.4 (d, J = 24.9 Hz), 18.8 (d,
J = 3.8 Hz), 18.5 (s), 10.0 (d, J = 13.4 Hz).
2.4. Preparation of PhP(o-C6H4–PEt2)2, 1
A 1.0 M THF solution of o-C6H4(PEt2)(I), 2i, (10.0 g, 34.25 mmol)
was cooled in an ice bath and slowly treated with 2.9 M i-PrMgBr
(11.81 mL, 34.25 mmol) in THF, then allowed to stir at 0 °C for 8 h.
This solution was then cooled to ꢁ25 °C by an acetone/dry ice bath
and slowly treated with a 1.0 M THF solution of PhPCl2 (3.064 g,
17.124 mmol), then allowed to warm to room temperature and
stirred for an additional 8 h. The solution was warmed to 70 °C
for 4 h and then allowed to cool to room temp. This solution was
quenched with 100 mL of an aqueous NH4Cl solution. The organic
layer was extracted and any remaining organic product was ex-
tracted from the aqueous layer with three 20 mL portions of ether.
The combined organic fractions were dried over Na2SO4 and a
gravity filtration through celite was performed. The solvent was re-
moved by vacuum evaporation and short path vacuum distillation
was used to remove any unreacted o-C6H4(PEt2)(I), 2i (0.5 torr, oil
bath of 155 °C). The remaining product can be recrystallized from
methanol or DMF. Typical isolated yields were 38–42%, although
31P NMR of the crude reaction mixture showed a yield of 70+%.
31P{1H} NMR (101.2 MHz, d ppm, C6D6): forms an AB2 second
order pattern that was simulated to give the two external phos-
phorus atoms at ꢁ26.2 ppm and the internal phosphorus at
ꢁ17.0 ppm with a calculated JP–P = 152.7 Hz (see Fig. 1). 1H NMR
(250 MHz, d ppm, C6D6): 7.5 (m, 2H), 7.4 (m, 2H), 7.2 (dd,
J = 14.1 Hz, J = 6.8 Hz, 7H), 7.0 (t, J = 7.3 Hz, 2H), 1.7 (dq, JH,H = 7.7 -
Hz, JH,P = 20.1 Hz, 8H), 1.02 (td, JH,P = 7.3 Hz, JH,H = 13.7 Hz, 12H.
13C{1H} NMR (62.8 MHz, d ppm, C6D6): 135.2 (d, J = 19.2 Hz),
134.7 (m), 130.5 (d, J = 7.7 Hz), 129.0 (d, J = 15.3 Hz), 20.6 (m),
10.4 (t, J = 7.6 Hz).
Mass spectral analyses were conducted at the LSU Mass Spec-
trometry Facility. ESI-MS was performed by an Agilent 6210 Elec-
trospray Time-of-Flight Mass Spectrometer. Samples of 3 and 4
were dissolved in DCM and analyzed in both positive and nega-
tive-ion mode.
2.1. Preparation of 1-(diethylphosphino)-2-iodobenzene, 2i
The following procedure was performed in a Schlenk flask cov-
ered with aluminum foil in order to exclude light. The iodo-prod-
ucts are light sensitive so it is important to protect them from
even fluorescent lab lights.
A solution of 1,2-diiodobenzene
(25.0 g, 75.77 mmol) in THF (80 mL) was treated at 0 °C with a
0.44 M THF solution of i-PrMgBr (171.0 mL, 75.77 mmol). The
resulting solution was stirred at 0 °C for 6 h. It was subsequently
cooled to ꢁ25 °C, and slowly treated with a solution of Et2PCl
(9.72 g, 78.0 mmol) in THF (90 mL). The yellow solution was al-
lowed to warm to room temperature and stirred overnight. Water
(80 mL) was added, and the organic layer was separated. The aque-
ous layer was extracted with diethyl ether (3 ꢂ 50 mL), the organic
extracts were combined and dried over Na2SO4, and the solvents
were removed in vacuo to yield a slightly yellow liquid. The prod-
uct was distilled via short-path distillation in vacuo to yield 16.0 g
(72%) of an air- and light-sensitive colorless liquid: bp = 116–
122 °C (0.6 Torr). Typical isolated yields are 70–75%, and the purity
of the product is typically greater than 99% based on NMR.
31P{1H} NMR (101.2 MHz, d ppm, C6D6): 0.3 (s). 1H NMR
(250 MHz, d ppm, C6D6): 7.7 (br m, 1H), 7.2 (sharp m, J = 7.3 Hz,
2H), 6.8 (sharp m, J = 7.3 Hz, 1H), 1.5 (m, 4H), and 0.9 (m,
2.5. Preparation of Ni(NCS)2[j
3-PhP(o-C6H4–PEt2)2], 3
In a Schlenk flask 0.879 g (2.00 mmol) of ligand 1 were dis-
solved in 70 mL of ethanol. Another Schlenk flask was charged with
0.682 g (2.00 mmol) Ni(BF4)2ꢀ6H2O and dissolved in 15 mL of eth-
anol. The ligand solution was added dropwise via cannula to the
Ni(BF4)2 solution. As the addition proceeded the solution became
very dark red in color. After the addition, a solution of 0.487 g
(5.00 mmol 2.5 equivalents) KSCN dissolved in 15 mL of ethanol
was added to the dark red solution. This solution was allowed to
stir overnight during which a red solid precipitated out of solution.
The next day the solid was collected via filtration and washed with
ethanol and diethyl ether and dried under vacuum. NMR analysis
revealed it to be the desired complex with a 65% isolated yield.
Dark red needle crystals for the X-ray analysis were grown by slow
evaporation of a THF solution.
J = 7.3 Hz, JP,H = 7.7 Hz, 6H). 13C{1H} NMR (62.8 MHz,
d ppm,
C6D6): 142.3 (d, J = 15.3 Hz), 139.5 (s), 139.4 (d, J = 15.3 Hz),
108.5 (d, J = 40.3 Hz), 77.4 (d, J = 30.7 Hz), 76.6 (s), 19.3 (s), and
9.5 (d, J = 13.4 Hz).
2.2. Preparation of 1-(diethylphosphino)-2-bromobenzene, 2b
The procedure described above was repeated on a 0.148 mol
scale using 1,2-dibromobenzene. The bromo products are not light
sensitive. The reaction afforded the product in 75–80% isolated
yield (28.3 g): bp = 83 °C (0.4 Torr, lit. bp = 111–112 °C/0.8 Torr
[1]).
31P{1H} NMR (101.2 MHz, d ppm, C6D6): ꢁ13.6 (s). 1H NMR
(250 MHz, d ppm, C6D6): 7.4 (br m, 1H), 7.1 (d, J = 4.5 Hz, 2H), 7.0
(m, 1H), 1.6 (m, J = 7.3 Hz, 4H), and 0.9 (m, J = 7.3 Hz, JP,H = 7.7 Hz,
6H). 13C{1H} NMR (62.8 MHz, d ppm, C6D6): 139.8 (d, J = 17.2 Hz),
134.5 (s), 131.4 (d, J = 28.8 Hz), 129.3 (s), 78.1 (d, J = 32.6 Hz),
77.4 (s), 19.1 (s), and 10.3 (d, J = 13.4 Hz).
31P{1H} NMR (101.2 MHz, d ppm, CD2Cl2): 59.2 (Pext, d, JP–
P = 56 Hz), 89.6 (Pint, t, JP–P = 56 Hz). 1H NMR (CD2Cl2): 0.95–1.04
(m, 6H), 1.27–1.36 (m, 6H), 2.08–2.28 (om, 6H), 2.45–2.55 (m,
2H), 6.67–6.72 (m, 2H), 7.32–7.36 (m, 2H), 7.44–7.49 (m, 1H),
7.73–7.87 (om, 8H).

W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
173
Fig. 1. 31P{1H} NMR spectrum of 1 (top) with AB2 simulation (bottom).
ESI-MS positive-ion mode m/z: found 554.0887 [MꢁSCNꢁ]+, cal-
culated 554.0899 for NiP3C27H33NS; negative-ion mode m/z: found
670.0406 [M+SCNꢁ]ꢁ, calculated 670.0403 for NiP3C29H33N3S3.
For ligand 1, one of the PEt2 groups was disordered into two
orientations with approximately 58:42 occupations. Distance
restraints and isotropic C atoms were necessary in refining this
disorder.
The crystal of 3 was twinned by 180° rotation about the real
[0ꢁ21] lattice direction, and the structure was refined versus
HKLF5 data, with refined BASF parameter 0.556(1). Of the six inde-
pendent complex molecules, one had a disordered ethyl group,
which was modeled with restraints similar to those used for 1.
The structure exhibits pronounced translational pseudosymme-
try. Transformation by the matrix (01/31/3, 02/3ꢁ1/3, ꢁ100)
produces an approximate cell having 1/3 the volume of the true
2.6. Preparation of NiCl2[j
3-PhP(o-C6H4–PEt2)2], 4
In a Schlenk flask 0.838 g (1.91 mmoles) of ligand 1 was dis-
solved in 70 mL of ethanol. Another Schlenk flask was charged with
0.454 g (1.91 mmoles) of NiCl2ꢀ6H2O and dissolved in about 20 mL
of ethanol. The ligand solution was added dropwise via cannula to
the NiCl2 solution. As the addition proceeded the solution changed
color from green to orange to dark red. After the addition, the solu-
tion was allowed to stir overnight. The next day, the solution was
evaporated to dryness under vacuum. The black/red residue was
redissolved in a minimum amount of dichloromethane and filtered
to remove any unreacted NiCl2. NMR analysis of the dichlorometh-
ane solution revealed the formation of the desired complex nearly
quantitatively. A large excess of hexanes was added to the concen-
trated dichloromethane solution and a red solid precipitated. The
solid was collected via filtration, washed with diethyl ether and
dried under vacuum. NMR analysis revealed it to be the desired
complex with a 75% isolated yield.
cell and monoclinic cell dimensions a = 12.145, b = 13.788,
0
c = 17.371 ÅA, b = 95.188°, and only two independent molecules in
the asymmetric unit. The structure can be solved in space group
P21/c in this subcell and refined to R = 0.054 (compared to 0.053
for the larger cell); however, much more disorder is necessarily
present in that model. Furthermore, the 2/3 of the data which
would be systematically absent in the monoclinic cell are clearly
nonzero, having mean intensity 4.78
r above zero. Thus, we are
confident that the reported triclinic structure is correct. Crystal
data and data collection and refinement parameters are given in
Table 1.
31P{1H} NMR (d ppm, CD2Cl2): 55.5 (Pext, d, JP–P = 49 Hz), 90.6
(Pint, t, JP–P = 49 Hz). 1H NMR (d ppm, CD2Cl2): 1.04–1.24 (om,
12H), 2.06–2.33 (om, 6H), 2.55–2.62 (m, 2H), 7.12–7.17 (m),
7.37–7.41 (m), 7.97–8.04 (b), 8.32–8.36 (m).
2.8. DFT calculations
ESI-MS positive-ion mode m/z: found 531.0838 [MꢁClꢁ]+, cal-
culated 531.0837 for NiP3C26H33Cl; negative-ion mode m/z: found
601.0233 [M+Clꢁ]ꢁ, calculated 601.0214 for NiP3C26H33Cl3.
GAUSSIAN 2009 was used with GAUSSVIEW 5 to perform molecular
orbital calculations on o-C6H4(X)(PMe2) (X = F, Br). The DFT meth-
od was used with the B3YLP hybrid functional and 6-311G(d,p) ba-
sis set functions on all atoms. The starting molecules were built in
GAUSSVIEW 5 and three different rotational conformations of the
PMe2 group were optimized. Two of the conformations had the
phosphorus lone pair in plane with the arene ring and oriented
syn and anti to the halogen. The third conformation had the phos-
phorus lone pair oriented approximately perpendicular to the
arene ring. This last conformation ended up having the lowest en-
ergy for the bromo system with one of the methyl groups approx-
imately in plane with the phenyl ring and the lone pair at an
approximate 40° angle to the phenyl ring. This conformation in
the fluoro-system had a similar structure and lone pair/methyl
2.7. X-ray experimental
X-ray diffraction measurements were performed on a Nonius
KappaCCD diffractometer equipped with an Oxford Cryostream
chiller and 0 graphite monochromated Mo
(k = 0.71073 ÅA). Structures were solved by direct methods, and
refinements were carried out by full matrix least-squares tech-
niques on F2, using SHELXL [11]. The hydrogen atoms were placed
in idealized positions.
K
a
radiation

174
W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
Table 1
Crystallographic data.
1
3
Formula
C
26H33P3
C28H33N2NiP3S2
CCDC 892495
613.30
CCDC deposition #
Molecular weight
Temperature (K)
Crystal system
CCDC 892494
438.43
90.0(5)
monoclinic
P21/c
12.076(2)
90.0(5)
triclinic
ꢀ
Space group
P1
0
17.3714(10)
18.3727(10)
27.933(2)
a (AÅ)
0
12.713(2)
16.896(3)
b (AÅ)
0
c (AÅ)
a
b (°)
c
V (AÅ3)
(°)
78.215(3)
85.488(3)
86.576(4)
8691.4(9)
110.723(10)
2426.1(7)
(°)
0
Z
4
12
1.406
1.000
Dcalc (g cmꢁ3
)
1.200
1.256
l
Mo Ka
(mmꢁ1
)
Crystal size (mm)
Color
0.30 ꢂ 0.27 ꢂ 0.17
0.17 ꢂ 0.13 ꢂ 0.13
colorless fragment
red needle
Tmin/Tmax
0.927/0.958
0.844/0.885
h (°)
2.5–30.0
2.5–30.1
Total/unique/observed reflections (Rint
)
55988/7064/5983 (0.023)
179431/81108/46326 (0.043)
R [F2 > 2
r
(F2)]a
0.071
0.179
1.091
269
0.053
0.142
1.036
1972
6
wR2 [F2 all reflections]b
S
Parameters
Restraints
15
Residual electron density
0
+1.72
+0.76
D
D
qmax (e ÅAꢁ3
)
0
ꢁ1.99
ꢁ0.76
qmin (e ÅAꢁ3
)
group orientation and was only 0.48 kcal higher than the lowest
energy conformer, which had the phosphorus lone pair oriented
anti to the fluoride and approximately in the arene plane. We used
these conformations for our comparisons of charges and MO’s.
along with optimized reaction conditions. Knochel and coworkers
originally developed the use of i-PrMgBr or i-Pr2Mg for the func-
tionalization of iodobenzenes [12,13]. Hartley [2] used the method
of Quin and Humphrey [14] based on o-C6H4(Br)(N„N)](BF4) to
prepare the PCl2 derivative in only 21% yield.
We have found that either the bromo- or iodo-arenes are suit-
able starting materials for this chemistry giving nearly quantitative
yields of the o-C6H4(X)(PEt2) (X = F, Br, I) products based on 31P
NMR of the crude reaction mixtures. Isolated yields are all in the
65–80% range. The X = I product has not been previously reported
and is light sensitive. Although we did not work with any of the
chloroarenes for this chemistry, they should work fine so long as
an ortho bromo or iodo group is also present. This chemistry can
3. Results and discussion
3.1. Synthesis of o-C6H4(X)(PEt2) (X = F, Br, I)
The dramatically improved synthesis of o-C6H4(X)(PEt2) (X = F,
Br, I) is shown in Scheme 1. The key advance over the original prep-
aration by Hart [1] is the use of i-PrMgBr instead of n-butyl lithium,
Scheme 1.

W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
175
It is important to note that the temperature of the Et2PCl addi-
tion is also critical as no product is observed when it is performed
at 0 °C. The best results are obtained when the addition of i-PrMgBr
is performed at 0 °C, which avoids crystallization of the aforemen-
tioned Grignard reagent as well as aryl Grignard product, followed
by stirring at 0 °C for 6 h to allow the reaction to proceed to com-
pletion, and finally, addition of Et2PCl at ꢁ25 °C. Although all this
synthetic work has focused on using Et2PCl, we believe that this
methodology will work just as well with most other R2PCl precur-
sors. Very bulky phosphine substituents like t-Bu, however, would
probably make this coupling difficult and lower the yields. Li and
i
-
coworkers have reported the synthesis of
2 with R = t-Bu
and X = Br in 50% yield using a Pd-catalyzed reaction between
HP(t-Bu)2 and o-bromoiodobenzene (3 day reaction at 80 °C) [15].
The trend in the 31P{1H} NMR spectra of the o-C6H4(X)(PEt2)
(X = F, Br, I) products is interesting with singlets at (C6D6 solvent):
d = 0.3 ppm (X = I), ꢁ13.6 ppm (X = Br), and a doublet centered at
d = ꢁ23.1 ppm (X = F, JP,F = 37.3 Hz). The increasing electronegativ-
Scheme 2.
easily be extended to almost any R2PCl phosphine, although we
have only directly worked with the diethyl phosphine.
When o-iodophenylbromide was used as the starting material,
the formation of o-C6H4(Br)(PEt2) was highly dependent on the
addition temperature. Thus, the isolated yield went from 65%
(90+% based on 31P NMR of the crude reaction mixture) when both
additions were performed at ꢁ40 and ꢁ78 °C, respectively, to only
40% when both additions were performed at ꢁ25 °C and the aryl
Grignard reagent was allowed to warm up to ambient temperature.
The reactivity of the o-C6H4Br2 with i-PrMgBr is also highly
dependent upon temperature. The lack of reactivity of o-dibromo-
benzene at ꢁ78 °C led us to investigate higher reaction tempera-
tures. Performing the additions at 25 and 0 °C did not generate
any product, but performing the additions at 0 and ꢁ25 °C, respec-
tively, gave much better results. Allowing the aryl Grignard reagent
to warm to ambient temperature for extended periods of time
prior to the addition of PEt2Cl led to a decrease in the isolated reac-
tion yield from 43% to 10%.
ity (v) of the halogen ortho to the PEt2 phosphine group causes an
upfield shift (i.e., toward more negative chemical shifts), which
typically indicates a more electron-rich phosphorus center. This
particular trend appears unexpected at first, as the increase in elec-
tronegativity should withdraw electron density from the aryl ring
and phosphorus, which should shift the 31P signals downfield
(i.e., to more positive chemical shifts).
This effect is overridden by a
action of the lower energy lone electron pairs on the F nucleus with
the filled molecular orbitals on the aryl ring results in the desta-
bilization of the benzene -system due to filled–filled orbital
repulsions. This, in turn, results in an increase in the -donating
ability of the aryl conjugated system to the P nucleus. The
bromide and iodide filled p-orbitals, however, are higher in energy
p-pushing mechanism: the inter-
p
p
p
p
Fig. 2. ORTEP plot of PhP[o-C6H4(PEt2)]2, 1, with 50% probability ellipsoids. Only one of the disordered PEt2 groups at P3 is shown for clarity.

176
W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
relative to the aryl p-system and this causes a stabilization of the
p-system, resulting in less donation to the phosphine.
DFT calculations performed on o-C6H4(X)(PMe2) (X = F, Br) sup-
port this electronic effect. The fluoro-substituted system has a
Mulliken charge on the phosphorus atom of +0.483, while the bro-
mo-substituted compound has a slightly higher charge of +0.496.
The MO’s for the bromo-system shows that the lone pair MO’s that
are interacting with the arene
p-system are quite a bit higher in
energy relative to the fluorine analogs (Br: ꢁ6.86 and ꢁ8.65 eV
versus F: ꢁ7.02 and ꢁ12.80 eV) and stabilize several key arene
p-orbitals, resulting in less p-donation to the phosphine.
Tunney and Stille [16] have reported the synthesis of o-C6H4
(I)(PPh2) and o-C6H4(Br)(PPh2) via Pd-catalyzed P–C coupling reac-
tions. Their observed 31P{1H} NMR spectra show the same upfield
shift of the signal from 9.1 ppm (iodo) to ꢁ4.4 ppm (bromo), fitting
in well with our observations for the –PEt2 derivatives. They re-
ported 82% yield of o-C6H4(Br)(PPh2) via the Pd-catalyzed
(2.5 mol% catalyst) coupling chemistry at 70 °C over 100 h, but
only 20% yield of o-C6H4(I)(PPh2) at 60 °C over 73 h. Their chemis-
try is limited to iodo-substituted phenyl rings, while the current
i-PrMgBr chemistry works extremely well with bromo-substituted
systems.
3.2. Synthesis and characterization of the trisphosphine, PhP[o-
C6H4(PEt2)]2, 1
The synthesis of PhP[o-C6H4(PEt2)]2, 1, is shown in Scheme 2
and the second-order AB2 31P{1H} NMR spectrum in Fig. 1 with
the simulated 31P{1H} NMR spectrum in red. Although the isolated
yield of 1 is only around 40%, it is far better than what has been re-
ported previously when one takes into account the much improved
synthesis and yield on 2i, the key starting material for this
preparation.
The structure of 1 is shown in Fig. 2, in which only one of the
disordered PEt2 groups at P3 is illustrated. In the crystal, the mol-
ecule adopts a conformation in which the two benzoP2 groups are
roughly perpendicular, forming a dihedral angle of 81.77(2)° for
the major (58%) conformer and 88.42(3)° for the minor conformer.
The P-phenylene bond distances are very similar and range from
1.847(2) to 1.854(2) Å, with the central P1-phenyl distance a bit
shorter at 1.840(2) Å. There are no structures reported on similar
trisphosphine ligands, the closest related system is the o-pheny-
lene bridged tetraphosphine, P[o-C6H4–PMe2]3 [17]. The P-pheny-
lene bond distances for this tetraphosphine ligand are all shorter
and range from 1.813 to 1.823 Å. The three phenylene rings in this
ligand are also all oriented approximately perpendicular to one
another (85–88°).
3.3. Synthesis and characterization of Ni(X)2[j
3-PhP(o-C6H4(PEt2)2)],
(X = NCSꢁ,3; Clꢁ, 4)
Fig. 3. ORTEP plot of one molecule of Ni(NCS)2[j
3-PhP(o-C6H4(PEt2)2], 3, with 50%
probability ellipsoids (top). Least squares fit overlay of all six molecules in the
asymmetric unit based on the Ni and five ligand atoms coordinated to the Ni center
(bottom).
The reaction of one equivalent of Ni(BF4)2 with PhP[o-C6H4
(PEt2)]2, 1, followed by reaction with 2.5 equivalents of KNCS
proceeds in good yield (65% isolated) to produce red Ni(NCS)2
[PhP(o-C6H4(PEt2)2)], 3. The structure of one of the six molecules
of 3 in the asymmetric unit is shown in Fig. 3. Also shown is a least
squares fit (based on the Ni center and five ligand atoms coordi-
nated to the Ni) and overlay of all six molecules in the asymmetric
unit. The rms deviation of the NiP3N2 group over the six molecules
is 0.062 Å. The conformations of the six molecules are, therefore,
quite similar with what we consider to be the expected variations
in the somewhat more flexible ligand periphery. There is some
variability in the PEt2 group at P3, including disorder in one of
the molecules. Due to the similarly of the core structures, average
bond distances and angles are reported in Table 2.
Structures with Z0 = 6 are unusual. A search of the Cambridge
Structural Database (version 5.33, May 2012 update) yielded only
221 hits, of which six have been reinterpreted with higher symme-
try and lower Z0. Most, 148, were determined below room temper-
ꢀ
ature, and a good number, 69, are in space group P1. Our
experience with phase changes to lower symmetry structures on
cooling suggests that the structure of this Ni complex may revert
to the Z0 = 2, P21/c space group, to which it is pseudosymmetric,
at higher temperatures.

W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
177
trisphosphine, EtP(CH2CH2PPh2)2 [20], 5, that is an Pint-ethyl-
substituted analog of the classic triphos ligand. This complex
adopts a structure much closer to trigonal bipyramidal. For exam-
ple, the P–Ni–P angle in 3 averages 152°, while in 5 it is 137°. The
difference in coordination environments is also reflected in the
trans-like P–Ni–N angle of 165° in 3 relative to 174° in 5. The
Ni–P2,3 distances in 5 are 2.19 and 2.22 Å, with the internal
Ni–P1 distance compressed at 2.15 Å. The Ni–P distances in 3 are
a bit shorter: Ni–P2,3 = 2.19, 2.20 Å and Ni–P1 = 2.13 Å, which is
tied into the stronger chelate effect of this ligand.
The reaction of one equivalent of NiCl2ꢀ6H2O with PhP[o-C6H4
(PEt2)]2, 1, proceeds quantitatively (as monitored by NMR) to
produce red NiCl2[PhP(o-C6H4(PEt2)2)], 4. This was characterized
by NMR and has similar 31P and 1H spectra to the thiocyanate
complex 3.
Table 2
Selected geometric parameters for Ni(NCS)2[j
3-PhP(o-C6H4(PEt2))2], 3.
Atoms
Distance (Å) or angle (°) range
Mean value (of 6)
Ni1–P1
Ni1–P2
Ni1–P3
Ni1–N1
2.1270(12)–2.1316(12)
2.1920(12)–2.2054(13)
2.1815(13)–2.1966(13)
1.895(3)–1.907(3)
2.081(4)–2.105(4)
84.53(5)–85.87(4)
2.129
2.200
2.186
1.900
2.093
85.2
Ni1–N2
P1–Ni1–P2
P1–Ni1–P3
P1–Ni1–N1
P1–Ni1–N2
P2–Ni1–P3
P2–Ni1–N1
P2–Ni1–N2
P3–Ni1–N1
P3–Ni1–N2
N1–Ni1–N2
Ni1–N1–C27
Ni1–N2–C28
86.81(5)–87.19(5)
87.0
164.57(13)–167.04(13)
98.59(10)–100.00(10)
147.63(6)–156.62(6)
91.42(11)–92.88(11)
101.24(11)–103.59(11)
88.25(11)–90.26(11)
101.00(11)–110.03(11)
93.76(15)–95.65(15)
166.9(4)–175.1(4)
166.2
99.5
152.2
92.2
102.2
89.1
105.4
94.4
4. Conclusion
171.2
166.1
163.6(3)–169.6(3)
A dramatically improved synthesis of o-C6H4(X)(PEt2) (X = F, Br,
I) from o-C6H4(X)(X0) (X = Br, I; X0 = F, Br, I) and Et2PCl using
i-PrMgBr is reported that gives quantitative crude yields and iso-
lated yields around 75%. o-C6H4(X)(PEt2) is the key precursor for
preparing the strongly chelating trisphosphine PhP[o-C6H4(PEt2)2],
1. Although we only worked with PEt2, the synthesis should be
general for a wide variety of other substituted phosphines. 1 was
characterized by NMR, crystallized, and structure confirmed by
X-ray crystallography. The reaction of Ni(BF4)2 and KNCS or
NiCl2ꢀ6H2O with one equivalent of
1 in produces Ni(X)2
[j
3-PhP(o-C6H4(PEt2))2], (X = NCS, 3; Cl 4) in high yields. Both com-
plexes have been characterized by NMR along with a single crystal
structure on 3. The structure is distorted square pyramidal with six
molecules of 3 in the asymmetric unit. All six molecules of 3 have
similar structures with minor conformational changes in the ligand
periphery.
Acknowledgements
The authors are grateful to NSF and Sasol North America for
funding that partially supported this project.
Appendix A. Supplementary material
Fig. 4. Rich Eisenberg and Michelle Millar at the 2009 Inorganic Chemistry Gordon
Research Conference (her last one).
CCDC 892494 and 892495 contain the supplementary crystallo-
graphic data for PhP(o-C6H4(PEt2))2, 1, and Ni(NCS)2[j
3-PhP(o-C6
H4(PEt2))2], 3, respectively. These data can be obtained free of
charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or
from the Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336 033; or e-mail:
deposit@ccdc.cam.ac.uk.
The coordination geometry of the Ni center is square pyramidal,
with one N-coordinated thiocyanate in a basal position and the
other at the apical position. The four basal ligand atoms are copla-
nar to within a mean deviation of 0.121 Å, and the Ni atom lies
0.381 Å out of this plane, both values averaged over the six mole-
cules. The structure of the free ligand, 1, had dihedral angles be-
tween the two benzoP2 groups of 81.77(2)° and 88.42(3)° for the
two disordered conformers. On complexation to Ni, this dihedral
angle drops to 42.0°, averaged over the six independent molecules
in the asymmetric unit.
Table 2 gives the details of the coordination geometry, showing
both the range of values for each distance and angle, as well as the
mean of the six values for each. The central Ni–P1 distance is seen
to be about 0.06 Å shorter than the external Ni–P2,3 distances, and
the bite angles of both chelate rings are near 86°. The axial Ni–N
distance is, as expected, 0.19 Å longer than the basal NCS ligand.
Both thiocyanates are seen to deviate slightly from linear coordina-
tion, by amounts up to about 16°, typically around 11°.
Appendix B. In memorial
I (GGS) first met Michelle Millar in August of 1975 as a new
graduate student in Al Cotton’s group. Michelle and her husband
Steven Koch were both postdoctoral fellows in the lab next to
mine. She was a bright and shining light in the lab from both a
chemistry and personality viewpoint. Michelle and Steve are two
of the best synthetic inorganic–organometallic chemists I have
ever had the pleasure of knowing.
I
vividly remember when Michelle first synthesized
[W2(CH3)8]4ꢁ after months of effort (no, Wilkinson did NOT make
it first, Michelle did). It was pyrophoric and decomposed above
ꢁ20 °C. The crystals had to be mounted in a glove bag with a ste-
reomicroscope in a ꢁ30 °C freezer room. A number of us in the
group took turns working in the freezer room to try and get a
crystal mounted in a capillary, but Michelle was by far the most
There are a couple of Ni(II) structures based on the phenylated
version of 1 [18,19]. But, there is only one other structurally
characterized Ni(NCS)2(P3) system based on the more flexible

178
W.J. Schreiter et al. / Polyhedron 58 (2013) 171–178
persistent and finally succeeded. The capillary was held above a
container of liquid nitrogen to keep it cool while transporting it
down the hall to the diffractometer equipped with a low tempera-
ture system.
things with an optimistic attitude. She was a passionate supporter
of junior faculty (myself included) and did yeoman service for her
university, the ACS (program chair for the Division of Inorganic
Chemistry for a number of years), and the Gordon Research Confer-
ences – especially Inorganic Chemistry (Fig. 4).
Once data collection started Michelle informed Al Cotton who
was extremely excited and wanted to see the crystal. But Texas
summer humidity created a frost shield around the low tempera-
ture nitrogen stream that obscured the crystal. Al reached over
and brushed the frost away, but misjudged and knocked the capil-
lary and crystal off the goniometer. ‘‘Ooops’’, I’m pretty sure is an
exact quote from Al, but the look of horror on Michelle’s face
was what I most certainly remember. Michelle could be impulsive
and for a few seconds she had a look indicating that Al was not long
for this world. Fortunately, she controlled herself and got another
mounted crystal from the freezer room, which turned out to dif-
fract even better than the first.
Michelle and Steve were also the driving synthetic (& crystallo-
graphic) forces in discovering the ‘‘supershort’’ Cr–Cr quadruply
bonded family of compounds at this same time. Yes, these were
heady times in Al’s group with Michelle and Steve constantly in
the middle of almost all the big discoveries.
Michelle, we will certainly miss you!
References
[1] F.A. Hart, J. Chem. Soc. (1960) 3324.
[2] J.G. Hartley, L.M. Venanzi, D.C. Goodall, J. Chem. Soc. (1963) 3930.
[3] J. Chatt, H.R. Watson, J. Chem. Soc. (1961) 4980.
[4] J. Chatt, F.A. Hart, H.R. Watson, J. Chem. Soc. (1962) 2537.
[5] J. Chatt, F.A. Hart, J. Chem. Soc. (1965) 812.
[6] M.A. Bennett, S.K. Bhargava, D.C.R. Hockless, L.L. Welling, A.C. Willis, J. Am.
Chem. Soc. 118 (1996) 10469.
[7] C.M. Conifer, D.J. Law, G.J. Sunley, A.J.P. White, G.J.P. Britovsek, Organometallics
30 (2011) 4060.
[8] K.D. Conroy, W.E. Piers, M. Parvez, J. Organomet. Chem. 693 (2008) 834.
[9] R.C. Taylor, R. Kolodny, D.B. Walters, Synth. React. Inorg. Met.-Org. Chem. 3
(1973) 175.
[10] B.E. Love, E.G. Jones, J. Org. Chem. 64 (1999) 3755.
[11] G.M. Sheldrick, Acta Crystallogr., Sect. A 64 (2008) 112.
[12] L. Boymond, M. Rottlander, G. Cahiez, P. Knochel, Angew. Chem., Int. Ed. 37
(1998) 1701.
[13] N. Gommermann, C. Koradin, P. Knochel, Synthesis 14 (2002) 2143.
[14] L.D. Quinn, S. Humphrey, J. Am. Chem. Soc. 83 (1961) 4124.
[15] Y.-H. Li, Y. Zhang, X.-H. Ding, Inorg. Chem. Commun. 14 (2011) 1306.
[16] S.E. Tunney, J.K. Stille, J. Org. Chem. 52 (1987) 748.
[17] D. Werner, L. Dahlenburg, Z. Naturforsch., B 42 (1987) 1110.
[18] F. Bigoli, S. Curreli, P. Deplano, L. Leoni, M.L. Mercuri, M.A. Pellinghelli, A. Serpe,
E.F. Trogu, J. Chem. Soc., Dalton Trans. (2002) 1985.
[19] F. Bigoli, P. Deplano, M.L. Mercuri, M.A. Pellinghelli, G. Pintus, E.F. Trogu, Chem.
Commun. (1999) 2093.
[20] B. Juma, F.R. Fronczek, G.G. Stanley, Acta Crystallogr., Sect. C 49 (1993) 1325.
Michelle was always extremely nice to me, even though both
she and Steve were somewhat dismayed when Al directed me into
computational studies (synthesis is ‘‘real chemistry’’ in both our
books). We lost touch during my postdoctoral years overseas, but
reconnected when I started my career in academics. Despite having
a number of personal setbacks Michelle always kept a positive out-
look. I was fortunate to know her well enough that she would not
hesitate to open up to me in private, but always ended up viewing